Field test of the TOXSWA pesticide fate model : Comparison of simulated and observed chlorpyrifos in water, sediment and macrophytes in four stagnant ditches

Samenvatting

TOXSWA is a numerical model describing pesticide behavior in an edge-of-field waterbody. It is widely used to predict exposure in regulatory risk assessment for aquatic ecosystems. Exposure concentrations are predicted based upon pesticide process parameters obtained in standardized laboratory experiments. However, few tests of the model performance based on field data have been carried out. We compare simulated concentrations to observations from a field experiment with four shallow stagnant ditches over sprayed with chlorpyrifos, a moderately volatile pesticide with a significant sorption capacity. Input parameters describing the four ditches, such as dimensions, water depth, sediment and macrophyte characteristics were measured in detail. Additionally, laboratory experiments were carried out to determine site-specific values for parameters describing chlorpyrifos degradation in water and sediment, as well as sorption to the two dominant macrophyte species. Based upon these estimated parameters, the correspondence between simulated and measured concentrations in water, sediment and macrophytes is poor. We attribute this discrepancy to a lack of site-specific input for the processes of volatilization and sorption to sediment, which both are important processes for chlorpyrifos. Therefore, we calibrated TOXSWA using the optimization tool PEST. The transfer coefficient for volatilization and the coefficient for sorption to sediment were optimized based on the observed concentrations in water and sediment. This resulted in a substantial improvement of correspondence. Optimized values of the transfer coefficient for volatilization and the coefficient for sorption to sediment are substantially higher than their initial estimates (4–8-fold and 2–4-fold increase, respectively), but can be well explained. The optimized coefficients vary less than a factor 2 between the four ditches. We conclude that TOXSWA can adequately predict chlorpyrifos behavior in the four ditches, provided that reliable site-specific parameter estimates are available. Field tests for other pesticides, waterbodies and agro-environmental conditions are warranted.